List your company

Embodied Referring Expression Comprehension on CAESAR-PRO

42.91IoU (0.25)CLIP
24.980429.635234.2938.9448Dec 6, 2025

Evaluation results

MethodLinks
CLIPResidual Module Configuration=MuRes(V)2025.1242.9111.91
Dual-EncoderResidual Module Configuration=Vanilla Residual2025.1242.6111.61
Dual-EncoderResidual Module Configuration=Without Residual2025.1242.5212.14
CLIPResidual Module Configuration=MuRes(L)2025.1239.5610.85
CLIPResidual Module Configuration=Vanilla Residual2025.1239.4310.83
CLIPResidual Module Configuration=MuRes(V+L)2025.1239.0610.46
Dual-EncoderResidual Module Configuration=MuRes(L)2025.1237.9710.32
CLIPResidual Module Configuration=Without Residual2025.1237.929.82
Dual-EncoderResidual Module Configuration=MuRes(V+L)2025.1237.7211.5
Dual-EncoderResidual Module Configuration=MuRes(V)2025.1236.728.51
ViLTResidual Module Configuration=MuRes(V+L)2025.1228.528.04
ViLTResidual Module Configuration=Without Residual2025.1227.968.73
ViLTResidual Module Configuration=Vanilla Residual2025.1225.678.06